School Choice: It’s Easy as 1, 2, 3, but Not for IDOE
“I tell yah Dave, I’ve been over this stuff a bunch of times; it just doesn’t add up. Who does these books? I mean if I ran my business this way I’d be out of business.”
That line from the movie “Dave” perhaps best sums up my review of the Indiana Department of Education’s (IDOE) calculation of the state’s Choice Scholarship Program—showing a near $16 million net loss for the school year just completed.
Although IDOE’s calculation may not be technically wrong, as they claim they’ve followed the formula specified in state law, it’s certainly grossly misleading. Using IDOE’s data and a little common sense, I can show how the true fiscal effect of Indiana’s Choice Scholarship Program is more in the range of a $16 million net savings.
The heart of the problem is the state-prescribed accounting formula, which is not designed to measure the true economic effect of school choice. Instead, its sole purpose is to ensure that, if school choice induces kids to switch from public school to private school faster than the state has projected, all incremental savings are redistributed back to the public schools. (Sadly, after examining the formula more closely, I’m not even certain it achieves that narrow goal.)
What’s quietly perverting IDOE’s calculation is a subtle (but critically important) mischaracterization of which voucher students generate a savings and which do not. The fundamental metric for calculating an accurate estimate of school choice savings is a breakdown between which voucher students were “diverted” from public school versus those “not diverted,” meaning:
- “Diverted” student: Any student that enrolls in a private school instead of a public school using a voucher. If the voucher is worth less than the per-pupil amount the state awards to public schools, the state saves money.
- “Not diverted” student: Any voucher student enrolled in private school who still could have enrolled without receiving a voucher. These students are, in fact, an extra cost to the state.
- “Previously enrolled in public school”: IDOE’s proxy for “diverted” students.
- “Not previously enrolled in public school”: IDOE’s proxy for “not diverted” students.
That distinction sound reasonable, but it’s not: Many of the students “not previously enrolled in public school” are actually “diverted” students who save the state money. Here’s why:
The most obvious flaw in the IDOE breakdown is that all voucher students entering kindergarten or first grade are counted as a net cost to the state. Because they were never enrolled in a public school, their calculation treats all of those kids as “not diverted” from public school. But history shows why that label is inaccurate.
Before there was ever a voucher program in Indiana, about 93 percent of all K-12 students were enrolled in public schools—just 7 percent were accessing private or homeschool environments. Furthermore, because Indiana’s school choice programs target low- to middle-income families, common sense dictates those families most likely would have to use public schools because they can’t afford private school tuition without school choice’s financial assistance.
With 19,811 students participating in the voucher program during the 2013-14 school year, to generate a $16 million net loss as the IDOE did, its calculation must assume that about 46 percent of the voucher students would still be able to cover their private school tuition without financial aid—46 percent! Here’s how IDOE must have reached that number:
- IDOE reports that the total cost of vouchers for the 19,811 students was about $80 million.
- To get a $16 million net loss, IDOE must have estimated that only $64 million was saved from students diverted from public schools ($64 million savings in aid to public schools – $80 million voucher costs = -$16 million net loss).
- Because the average per-pupil amount the state awards public schools is about $6,000, a $64 million savings in state aid to public schools represents about 10,700 students “diverted” from public schools (or 54 percent of the voucher students).
- Thus, the other 46 percent (all from low- to middle-income families) are presumed to be able to pay for their tuition without assistance.
I don’t know what actual numbers IDOE used in its calculation because those details were not provided when the results were reported—despite the lone public statement from IDOE’s spokesman that “(t)he Department is committed to a transparent accounting of how taxpayer dollars are spent.”
In January, when IDOE first provided a breakdown of voucher recipients, it reported that 7,779, or 39 percent, of the voucher students were “not previously enrolled in public school.”
How that count has since jumped up to 46 percent is something IDOE needs to explain.
But even back then I questioned IDOE’s use of the “previously enrolled” versus “not previously enrolled” breakdown suggesting instead that “diverted” versus “not diverted” was the proper measure.
Going a step further, I estimated the “diverted/not diverted” split, using IDOE’s own data, to be 81/19 percent. In other words, my analysis (executed with abundant caution) suggested that if the vouchers were pulled from all students, 19 percent would still figure out some way to pay their private school tuition without financial assistance. That estimate for the “not diverted” kids is far more plausible than IDOE’s 46 percent.
If my 81/19 split is in the ballpark, the fiscal impact of the Choice Scholarship Program flips from a near $16 million net cost to a $16 million net savings. That suggests more than 16,000 of the voucher students have been diverted from public schools. At $6,000 per pupil, the state is saving $96 million in aid to public schools. Comparing that to the $80 million state cost for the vouchers provides an entirely different and more reasonable result.
Ultimately, public education officials should dig more deeply into the existing redistribution formula to ensure it is delivering on its narrow goal of reallocating all unplanned school choice savings. If policymakers and other state officials want an accurate accounting of the true fiscal effect of school choice in Indiana, they’ll need to design an entirely different formula for measuring its costs and savings.
That’s a task with which I’m certainly ready to help.